Early extensional detachments in a contractional orogen · Draft Early extensional detachments in a contractional orogen: coherent, map-scale, submarine slides (mass transport complexes)

Draft

Early extensional detachments in a contractional orogen:

coherent, map-scale, submarine slides (mass transport complexes) on the outer slope of an Ediacaran collisional

foredeep, eastern Kaoko belt, Namibia

Journal: Canadian Journal of Earth Sciences

Manuscript ID cjes-2015-0164.R1

Manuscript Type: Article

Date Submitted by the Author: n/a

Complete List of Authors: Hoffman, Paul; 1216 Montrose Ave Bellefroid, Eric; Yale University, Geology and Geophysics Johnson, Benjamin; University of Victoria, School of Earth and Ocean Sciences Hodgskiss, Malcolm ; McGill University, Earth and Planetary Sciences Schrag, Daniel; Harvard University, Earth and Planetary Sciences Halverson, Galen; McGill University, Earth and Planetary Sciences

Keyword: submarine landslide, mass transport complex, extensional detachment,

slope failure, continental margin

https://mc06.manuscriptcentral.com/cjes-pubs

Canadian Journal of Earth Sciences

Draft

cjes-‐2015-‐0164 (revised, 12 Jan 16) 1

2

CJES Special Issue: Uniformitarianism and Plate Tectonics 3

A Tribute to Kevin C. Burke and John F. Dewey 4

5

6

Early extensional detachments in a contractional orogen: coherent, 7

map-‐scale, submarine slides (mass transport complexes) on the outer 8

slope of an Ediacaran collisional foredeep, eastern Kaoko belt, Namibia 9

10

11

Paul F. Hoffman1, Eric J. Bellefroid2, Benjamin W. Johnson3, 12

Malcolm S. W. Hodgskiss4, Daniel P. Schrag5 and Galen P. Halverson4 13 14

11216 Montrose Avenue, Victoria, BC V8T 2K4 15 2Department of Geology and Geophysics, Yale University, New Haven, CT 06520-‐8109 16

3School of Earth and Ocean Sciences, University of Victoria, Victoria, BC V8P 5C2 17 4Department of Earth and Planetary Sciences, McGill University, Montreal, QC H3A 0E8 18

5Department of Earth and Planetary Sciences, Harvard University, Cambridge, MA 02138 19

20

Corresponding author: [email protected] 21

Page 1 of 35



Draft

2

2

Abstract: The existence of coherent, large-‐scale, submarine landslides on modern 22

continental margins implies that their apparent rarity in ancient orogenic belts is 23

due to non-‐recognition. Two map-‐scale, coherent, pre-‐orogenic, normal-‐sense 24

detachment structures of Ediacaran age are present in the Kaoko belt, a well-‐25

exposed arc-‐continent collision zone in northwestern Namibia. The structures occur 26

within the Otavi Group, a Neoproterozoic carbonate shelf succession. They are 27

brittle structures, evident only through stratigraphic omissions of 400 m or more, 28

that ramp down to the west with overall ramp angles of 1.1° and 1.3° with respect to 29

stratigraphic horizons. The separations of matching footwall and hangingwall 30

stratigraphic cut-‐offs require horizontal translations >20 km for each detachment. 31

One of the detachments is remarkably narrow (5 km) in the up-‐dip direction, just 32

one fourth of its translation. The other detachment is stratigraphically dated at the 33

shelf-‐foredeep transition, when the passive margin was abortively subducted 34

westward, in the direction of submarine sliding. Trenchward sliding on the 35

foreslope occurred concurrently with deep karstification of the autochthonous 36

carbonate succession to the east, presumably due to forebulge uplift and/or 37

conjectural basin-‐scale base-‐level fall. We expect that similar detachments exist in 38

other orogenic belts, and failure to recognize them can lead to misinterpretations of 39

stratigraphy, sedimentary facies and paleogeography. 40

41

Keywords: submarine landslide, mass transport complex, extensional detachment, 42

slope failure, continental margin, collisional foredeep 43

44

45

Page 2 of 35



Draft

3

3

Introduction 46

Coherent, large-‐scale, submarine landslides exist on many modern continental 47

margins (Dingle 1977; Bugge et al. 1987; Kayen and Lee 1991; Hampton et al. 1996; 48

Rowan et al. 2004; Frey-‐Martinez et al. 2005; Masson et al. 2006; Butler and Paton 49

2010; Morley et al. 2011; Armandita et al. 2015). Bathymetric and/or seismic 50

reflection studies show them to consist of up-‐dip extended domains and down-‐dip 51

shortened domains, characterized by breakaway-‐ and thrust-‐duplexes respectively 52

(Fig. 1). Although stratigraphically confined, the presence of growth strata (Fig. 1A) 53

show some examples to have developed slowly by creep. Their apparent rarity in 54

ancient orogenic belts is more likely due to non-‐recognition than non-‐occurrence. 55

The discovery of the Ombonde detachment (Hoffman and Hartz 1999) in the 56

Kaoko belt of Namibia (Fig. 2) provided an Ediacaran example for which the net 57

displacement, paleotectonic setting and longitudinal (i.e., slip-‐parallel) cross-‐section 58

of the extended domain (Fig. 1) are well constrained by plunge projection of surface 59

maps and stratigraphic logs (Hoffman and Hartz 1999; Hoffman and Halverson 60

2008). What the Ombonde detachment did not reveal was its transverse (i.e., slip-‐61

normal) dimension, and an implicit shortened domain (Fig. 1) in the down-‐dip 62

direction, which is obscured by post-‐orogenic cover. The shortened domain of a 63

possibly analogous detachment, the ‘Saturn slide’ (Clifford 1962, 2008), was 64

recognized earlier within correlative strata of the adjacent Outjo zone (Fig. 3). 65

Here we describe a second low-‐angle extensional detachment in the Kaoko belt 66

(Fig. 3). Like the Ombonde detachment, 90 km to the south, the Ombepera 67

detachment predates orogenic shortening. It is a brittle structure within a carbonate 68

platform succession, locally associated with a dilational breccia zone in the 69

hangingwall, but lacking recognizeable shear fabrics. The structure is only evident 70

through stratigraphic mapping. The average ramp angle between the detachment 71

and stratigraphic planes is 1.3°, similar to the Ombonde detachment (1.1°). The 72

characteristic stratigraphic omission is somewhat greater than the 395 m of missing 73

strata across the Ombonde detachment, and the separation of stratigraphic cut-‐offs 74

Page 3 of 35



Draft

4

4

in the footwall and hangingwall of the Ombepera detachment imply >20 km of 75

displacement. 76

The timing of the Ombonde detachment is tightly constrained. Movement 77

postdates the entire Otavi Group carbonate shelf sequence and predates the basal 78

foredeep clastics of the disconformably overlying Mulden Group (Fig. 4)(Hoffman 79

and Hartz 1999). We assume contemporaneity of the two detachments based on 80

structural homology. Like the Ombonde detachment, only the extended domain of 81

the Ombepera detachment is exposed. Unlike the Ombonde detachment, its 82

transverse extent is directly constrained by surface mapping. This, and mutual 83

verification of the phenomenon, make the Ombepera detachment a structure of 84

interest. Failure to recognize the detachment would have led to serious 85

misinterpretations of the stratigraphy and sedimentary facies relations. 86

We begin by describing the basic stratigraphy of Neoproterozoic cover on the 87

southwestern Congo craton and its relation to the Kaoko and Damara orogenic belts, 88

in the context of centre-‐west Gondwana amalgamation. Next we review the salient 89

map-‐scale features of the Ombonde detachment that bear on its geometry, 90

displacement and age. Then we describe how the Ombepera detachment came to be 91

recognized in its type area, and how its interpretation was tested and confirmed by 92

up-‐dip projection. Finally, we identify the need for additional mapping to investigate 93

marked widening and/or flattening of the Ombepera detachment surface in the 94

down-‐dip direction 95

96

Neoproterozoic cover of the southwestern Congo craton 97

Paleoproterozoic crust of the present southwestern promontory of the Congo 98

craton (Fig. 2) is blanketed by a Neoproterozoic sedimentary succession composed 99

of three groups (Fig. 4). The basement-‐hugging Nosib Group (SACS 1980) consists of 100

subfeldspathic arenite and rudite deposited on a southward-‐sloping alluvial 101

braidplain. The conformably overlying Otavi Group (SACS 1980) is a 2-‐ to 4-‐km-‐102

thick carbonate shelf sequence with a well-‐defined shelf-‐break and a distally-‐103

tapered southern foreslope (Hoffman and Halverson 2008). A presumed western 104

Page 4 of 35



Draft

5

5

shelf margin is occluded by the Sesfontein thrust (Fig. 3), an eastward-‐vergent, 105

crustal-‐scale, thrust fault carrying the Central Kaoko zone, where the Otavi Group is 106

an off-‐shelf facies. 107

Internally, the Otavi Group consists of three subgroups (Fig. 4), divided by 108

unconformities that underlie each of the two Cryogenian glacial-‐periglacial 109

assemblages (Hoffmann and Prave 1996). The Chuos and Ghaub formations 110

represent the prolonged Sturtian (717-‐659 Ma) and briefer Marinoan (ca 645-‐635 111

Ma) glaciations, respectively. The Rasthof and Maieberg formations represent the 112

respective postglacial cap-‐carbonate sequences (Hoffman and Halverson 2008). 113

The Tonian Ombombo Subgroup (Hoffmann and Prave 1996) consists of cyclical, 114

shallow-‐marine dolomite with authigenic chert, intercalated with alluvial and 115

deltaic clastic rocks. Clastic input resulted from intermittent uplift and erosion of 116

northward-‐inclined dip-‐slopes, interpreted as the footwalls of spaced, south-‐117

dipping, normal faults (Hoffman and Halverson 2008). The age of the Nosib-‐Otavi 118

transgression and the onset of rift faulting must be significantly older than 760 ± 1 119

Ma (Halverson et al. 2005), the U-‐Pb zircon date of a bentonite near the top of the 120

carbonate-‐rich Devede Formation (Fig. 4) in the middle Ombombo Subgroup. 121

Intermittent rift-‐shoulder uplift, inferred from low-‐angle unconformities and 122

cannibalistic sedimentation, persisted into the Cryogenian Abenab Subgroup (SACS 123

1980) up to the base of the Ombaatjie Formation (Fig. 4), which marks the rift-‐to-‐124

shelf transition (Hoffman and Halverson 2008) at an estimated 650 Ma. Thereafter, 125

the shelf aggraded conformably and preferentially relative to the foreslope and 126

basin in the Outjo zone (Fig. 3) until the demise of carbonate sedimentation at the 127

top of the Ediacaran Tsumeb Subgroup (SACS 1980), the thickest and most 128

continuous of the Otavi subgroups. 129

The Mulden Group (SACS 1980) paraconformably overlies the Otavi Group and 130

forms an upward-‐shoaling, marine to nonmarine sequence of compositionally-‐131

immature clastics, culminating with crossbedded arenites (Renosterberg 132

Formation) deposited by high-‐discharge southeastward-‐flowing rivers (Hoffman 133

and Halverson 2008). On the southwest flank of the Kamanjab inlier (Fig. 3), paleo-‐134

outliers of Otavi Group carbonate <900 m thick are buried paraconformably by 135

Page 5 of 35



Draft

6

6

Mulden Group clastics which, at their base, were deposited directly upon the 136

crystalline basement (Frets 1969). The karstic outliers shed dolomite-‐chert debris-‐137

fans and conglomerates into the more far-‐travelled Mulden Group detritus. Locally, 138

the present drainages (e.g. upper Huab River at 20°06'10"S, 14°36'04"E) have re-‐139

incised Mulden-‐age paleovalleys (Frets 1969). The Mulden Group is interpreted as a 140

trench-‐foredeep assemblage associated with abortive westward subduction of the 141

Congo margin beneath the Dom Feliciano–Ribeira arc (Fig. 2)(Oyhantçabal et al. 142

2009; Chemale et al. 2012). Alternatively, it was a retro-‐arc foreland basin 143

developed above an east-‐dipping subduction zone (Goscombe and Gray 2007, 2008; 144

Konopásek et al. 2014). In our view, the largely amagmatic development of the rifted 145

western Congo margin (Guj 1970; Stanistreet and Charlesworth 1999; Goscombe et 146

al. 2005; Miller 2008) is more consistent with the first alternative. There is no 147

evidence for subduction-‐related magmatism in the Central Kaoko zone (see below), 148

in contrast with the Western Kaoko zone (Fig. 3). 149

150

Kaoko belt 151

Neoproterozoic cover and basement of the southwestern Congo craton are 152

deformed in an arcuate fold belt that manifests crustal shortening in the mutually 153

perpendicular Kaoko and Damara orogenic belts (Fig. 2). In the Eastern Kaoko zone 154

(Fig. 3), a train of tight, thin-‐skinned folds and subordinate thrusts formed above a 155

décollement within the Beesvlakte Formation (Ombombo Subgroup). This thin-‐156

skinned, eastward-‐directed, thrust-‐fold belt was subsequently refolded by broader, 157

coaxial, basement-‐involved structures, a relationship best observed at the northern 158

plunge of the Kamanjab basement inlier (Fig. 3). The northward and southward 159

plunges characteristic of Kaoko belt structures are at least in part a product of late-‐160

stage refolding as a far-‐field consequence of collisions in the Damara belt. 161

The Central Kaoko zone (CKz, Fig. 3) underwent sinistral transpression, dynamic 162

metamorphism, and rapid exhumation at 580-‐570 Ma, creating a system of 163

eastward-‐vergent, basement-‐cored, thrust nappes with northwest-‐plunging 164

stretching lineations (Dürr and Dingeldey 1996; Goscombe et al. 2003; Goscombe 165

Page 6 of 35



Draft

7

7

and Gray 2008). Metamorphic P-‐T-‐t paths are clockwise with limited coeval granitic 166

magmatism (Goscombe et al. 2003; Will et al. 2004; Jung et al. 2014). The CKz is 167

juxtaposed against the Western Kaoko, or Coastal, zone (WKz), by a continuous 168

high-‐grade ultramylonite zone, the Purros suture (Fig. 3), having sinistral 169

subhorizontal shear-‐sense indicators (Goscombe et al. 2003; 2007; Konopásek et al. 170

2005; 2008). The WKz has a magmatic and metamorphic history distinct from the 171

CKz (Goscombe et al. 2005; Goscombe and Gray 2007; Konopásek et al. 2008) and is 172

taken to represent the leading edge of the Dom Feliciano–Ribeira (DF-‐R) magmatic 173

arc of Uruguay and southern Brazil (Fig. 2), against which the Kaoko belt was 174

juxtaposed prior to South Atlantic opening (Oyhantçabal et al. 2009, 2011; 175

Konopásek et al. 2014). Deformation and exhumation of the CKz at 580-‐570 Ma 176

(Goscombe et al. 2005; Jung et al. 2014) was broadly coeval with a transition from 177

subduction-‐related to slab-‐failure magmatism and exhumation in the DF-‐R arc 178

(Oyhantçabal et al. 2009, 2011; Faleiros et al. 2011; Tupinambá et al. 2012; Alves et 179

al. 2013; Heilbron et al. 2013). Closure of the Adamastor paleocean, situated at this 180

latitude between the rifted Congo margin and DF-‐R arc, occurred 20-‐30 Myr after 181

abortive continental subduction of the opposite (eastward) polarity in the West 182

Gondwana belt (Fig. 2)(Ganade de Araujo et al. 2014), and >10 Myr before final 183

closure of the Brasiliano paleoocean, between the DF-‐R arc and the Amazonia craton 184

(Fuck et al. 2008; Bandeira et al. 2012; McGee et al. 2015a, b). 185

186

Damara belt 187

The Damara belt (Fig. 2) is an intact collision zone, riddled with bodies of post-‐188

collisional syenogranite, separating the Congo and Kalahari cratons. It is a 189

compound belt with two basinal domains, the Outjo and Khomas zones (Fig. 3), 190

divided by a central microcontinent, the southern Central Zone (Miller 2008) or 191

Swakop terrane (Hoffmann 1987). Terminal collision in the Damara belt occurred 192

when northward subduction within the more southerly Khomas zone led to collision 193

around 550 Ma between the arc-‐bearing Swakop terrane and the north-‐facing 194

passive margin (Witvlei Group) of the Kalahari craton. The fossiliferous late 195

Page 7 of 35



Draft

8

8

Ediacaran and early Cambrian Nama Group of the Kalahari craton is a compound 196

foredeep related to collisions incurred by outward-‐dipping subduction in the 197

Damara belt to the north and the Gariep belt to the west (Fig. 2). 198

Miller (2008) and most previous workers consider the more northerly Outjo zone 199

to have been floored by stretched continental crust of the Congo craton, the Swakop 200

terrane (Fig. 3) being interpreted as a horst block forming an outer basement high at 201

the rifted continental margin. However, subduction rarely if ever initiates at a 202

passive margin (Cloetingh et al. 1982), and we therefore favour a more complex 203

development in which the Swakop terrane collided first with a passive Congo 204

margin via southward subduction within the Outjo zone. The Outjo zone is 205

coextensive with the largest-‐amplitude and longest-‐wavelength aeromagnetic 206

anomaly trough in the country. We postulate that northward subduction in the 207

Khomas zone initiated through subduction zone ‘flip’ (McKenzie 1969; Suppe 1984), 208

consequent to a Swakop-‐Congo collision. We suggest that the N-‐S shortening event 209

recently dated by 40Ar/39Ar at ~590 Ma in the Outjo zone (Lehmann et al. 2015) is a 210

manifestation of Swakop-‐Congo collision in the Damara belt, broadly synchronous 211

with Congo-‐DF-‐R arc collision in the Kaoko belt. 212

213

Significance of the Otavi–Mulden group transition 214

Assuming the Kaoko belt is an arc-‐continent collision zone, the Otavi–Mulden 215

group transition (Fig. 4) records the passage of the upper Otavi Group carbonate 216

shelf margin over the outer forebulge of a west-‐dipping subduction zone and its 217

descent on the outer slope of the trench. Diachronous (southward-‐younging) 218

closure of the Adamastor paleocean and growth of the resulting collisional orogen 219

(Stanistreet et al. 1991) complies with paleocurrent evidence for southeastward 220

directed fluvial sediment transport (Renosterberg Formation) on the Congo 221

foreland (Hoffman and Halverson 2008). Rapid exhumation of the Kaoko belt at 222

580-‐570 Ma (Goscombe et al. 2005; Jung et al. 2014) implies an age of ~585 Ma or 223

older for the shelf-‐to-‐foredeep transition in the southern Kunene Region (Fig. 3). 224

225

Page 8 of 35



Draft

9

9

Ombonde detachment 226

The Ombonde detachment (Hoffman and Hartz 1999) is exposed for a strike 227

length of 51 km (22 km east-‐west) in the south-‐plunging Grootberg syncline (Fig. 5), 228

a basement-‐involved structure in the autochthonous foreland of the Kaoko belt. The 229

east limb of the syncline is steeply-‐dipping (50-‐70°) but the west limb is gentler and 230

crests in a broad anticline before disappearing beneath Early Cretaceous traps 231

(Etendeka Group). Paleogeographically, the Otavi Group is situated on the Makalani 232

dip-‐slope (Fig. 5C), where rift-‐shoulder uplift and erosion caused progressive 233

southward truncation of the Ombombo Subgroup, the Nosib Group, and the Rasthof 234

and Gruis formations. The Gruis Formation undergoes a remarkable facies change, 235

from peri-‐littoral carbonate cycles in the north to alluvial fanglomerate of exclusive 236

basement derivation 20 km to the south. Where intersected by the Ombonde 237

detachment, the Gruis Formation is a rheologically weak facies characterized by 238

marlstone and fine-‐grained clastic rocks. Other rheologically weak units are 239

limestones of the lower Rasthof, lower Ombaatjie and lower Maieberg formations, 240

while strong units are dolomites of the upper Rasthof, upper Ombaatjie, upper 241

Maieberg and Elandshoek-‐Hüttenberg (upper Tsumeb) formations. The Mulden 242

Group is represented by compositionally immature sandstone with metre-‐scale 243

crossbedding, locally with a basal chert-‐dolomite conglomerate containing clasts of 244

Elandshoek Formation stromatolite. 245

The Ombonde detachment was first identified in 1994 by Tony Prave and Dawn 246

Sumner on the east limb of the syncline (location h, Fig. 5A) by virtue of the fact that 247

~400 m of well-‐known strata (Ombaatjie and Maieberg formations) are missing 248

across a brittle, bedding-‐parallel, mappable, fault surface. The fault was 249

subsequently mapped throughout the syncline, revealing a ramp-‐flat-‐ramp 250

geometry (Fig. 5B) that cuts down-‐section westward from the lower Elandshoek 251

Formation to granitic basement in the footwall, and from the upper Hüttenberg 252

Formation to the Rasthof Formation in the hangingwall. The detachment is 253

commonly a single fault, but a 2.5-‐km-‐long extensional horse occurs at the exposed 254

keel of the syncline (f, Fig. 5A) and smaller horses occur elsewhere. Fault-‐related 255

Page 9 of 35



Draft

10

10

vein systems are uncommon, but stratigraphic cut-‐offs (Fig. 5B) provide essential 256

constraints on the fault-‐plane geometry and net slip. 257

The western ramp is exposed over the crest of the broad anticline (a-‐c, Fig. 5A). 258

The fault ramps down-‐section westward from the lower Elandshoek to the Rasthof 259

Formation in the hangingwall, and from the Gruis Formation to the basement in the 260

footwall. The hangingwall cut-‐off of the topmost Maieberg Formation is exposed on 261

both limbs of a second-‐order anticlinal crossfold (c, Fig. 5A), across which the 262

projected cut-‐off strikes 357°, implying a dip direction of 267° at that location. A dip 263

direction between 260° and 280° is required by the distribution of cut-‐offs 264

throughout the syncline in order for the detachment to ramp continuously down-‐265

section in one direction. It is noteworthy that the hangingwall is not structurally 266

extended in the panel that parallels the inferred slip direction (a-‐c, Fig. 5A). 267

Stratigraphic cut-‐offs in the hangingwall of the western ramp (a-‐c, Fig. 5A) are 268

matched by cut-‐offs in the footwall of the eastern ramp (h-‐j, Fig. 5A). The horizontal 269

separations of the corresponding cut-‐offs are 16.0, 16.4 and 17.3 km for the top of 270

the Gruis, Ombaatjie and Maieberg formations, respectively. Corrected for an 271

estimated 25% synclinal shortening, the respective horizontal separations are 20.0, 272

20.5 and 21.6 km (Fig. 5B). Because stratigraphic thicknesses are known from 273

measured sections (Hoffman and Halverson 2008), ramp angles between the 274

detachment surface and stratigraphic horizons can be calculated. The ramp angle 275

averaged through the Rasthof Formation in the footwall is 14°, while the respective 276

ramp angles through the Ombaatjie and Maieberg formations are 10° and 9° in the 277

hangingwall, and 8° and 9.5° in the footwall. The uncorrected footwall ramp angle 278

averaged through the entire Abenab Subgroup (395 m thick), including the long flat 279

within the Gruis Formation, is a mere 1.3°, or 1.1° corrected for synclinal shortening. 280

Compaction is slight in most Otavi Group carbonates; only in limestone (Fig. 4) are 281

stylolites common. If post-‐detachment compaction were 20%, at a maximum, the 282

overall ramp angle would increase by 0.2°. 283

The age of the Ombonde detachment is tightly constrained. High on the eastern 284

ramp, the detachment carries uppermost Otavi Group in the hangingwall onto upper 285

Maieberg Formation in the footwall (j, Fig. 5A). Fault slip must therefore post-‐date 286

Page 10 of 35



Draft

11

11

the entire Otavi Group. The critical minimum age constraint comes from basal 287

Mulden Group (Renosterberg Formation) paleovalleys, the largest of which is 288

located on the west limb of the syncline near the axial trace (d-‐e, Fig. 5A). This 289

paleovalley trends northeast-‐southwest and is filled by crudely bedded chert-‐290

dolomite conglomerate that dips ~20° toward the southeast, structurally 291

conformable with the underlying Otavi Group. Stromatolite clasts derived from the 292

upper Tsumeb Subgroup are prominent in the conglomerate, the structural attitude 293

of which precludes a Karoo (Carboniferous to Early Cretaceous) or younger age. On 294

the east wall of the paleovalley, the conglomerate forms a buttress unconformity 295

against the hangingwall of the detachment. The axial drainage of the syncline covers 296

the contact between the paleovalley and the detachment (e, Fig. 5A). On the west 297

wall of the paleovalley, the basal conglomerate bevels the detachment without 298

disturbance and variably abuts the Elandshoek Formation of the hangingwall and 299

both the Gruis and Rasthof formations of the footwall (d, Fig. 5A). Close to where the 300

paleovalley truncates the detachment, a small window in the hangingwall exposes 301

the Gruis Formation of the footwall. Erosional incision and truncation of the 302

detachment fault by the basal Mulden Group conglomerate (Fig. 5B), which is 303

conformably overlain by sandstone of the Renosterberg Formation, demonstrate 304

that the detachment must be pre-‐Mulden as well as post-‐Otavi in age. 305

High on the eastern ramp, a set of small paleogullies at the base of the Mulden 306

Group incises the uppermost Otavi Group in the hangingwall of the detachment (j, 307

Fig. 5A). The floors of these paleogullies reach the footwall Maieberg Formation. The 308

paleogullies confirm a post-‐Otavi, pre-‐Mulden group age for the detachment. 309

Hoffman and Hartz (1999) inferred that the Ombonde detachment formed as a 310

submarine slide on the outer slope of a trench that became a collisional foredeep 311

when the descending plate passed from oceanic to continental as the Congo margin 312

was abortively subducted at ~590 Ma. They postulated that coherent sliding of the 313

low-‐angle detachment was facilitated by pore-‐fluid overpressure, resulting from a 314

Messinian-‐type sea-‐level fall in the foredeep. As evidence for base-‐level fall, they 315

cited deep karstic erosion beneath the Mulden Group in the southwest of the Otavi 316

platform, incision of the conglomerate-‐filled paleovalley across the detachment, and 317

Page 11 of 35



Draft

12

12

the fluvial paleoenvironment of the Mulden Group (Renosterberg Formation) in the 318

area of the detachment. 319

One troubling fact regarding the Ombonde detachment emerged only over time. 320

After its nature and age were established by mapping in 1996, seventeen field 321

seasons elapsed before another example of the phenomenon was found in the Kaoko 322

belt. Such structures should not occur singly. 323

324

Ombepera detachment 325

Ninety kilometres to the northwest of the Grootberg syncline, the internal part of 326

the thin-‐skinned Eastern Kaoko zone includes a train of three synclines (S1-‐3, Fig. 327

6). Reconnaissance mapping and stratigraphic logging identified a problem in the 328

central syncline. The distinctive ‘cap dolostone’ of the younger Cryogenian 329

(Marinoan) glaciation, the normally ever-‐present Keilberg Member at the base of the 330

Maieberg Formation (Fig. 4), could not be found. In contrast, the cap dolomite was 331

easily recognized in the eastern and western synclines, in the latter atop glacigenic 332

carbonate diamictite of the Ghaub Formation. The conundrum of the central 333

syncline (S2, Fig. 6) remained unresolved until the area was revisited in the last 334

week of the 2014 field season. 335

Measuring the entire succession in the central syncline prompted a postulate 336

(Fig. 7) that the Gruis, Ombaatjie and lower Maieberg formations are missing across 337

a cryptic, normal-‐sense detachment that predates folding and thrusting, analogous 338

to the Ombonde detachment. The postulate was corroborated by carbon-‐isotope 339

chemostratigraphy. Strata directly overlying the Rasthof Formation are isotopically 340

depleted, δ13C = –2 to –5‰ VPDB (Section 2, Fig. 7). This is incompatible with the 341

Gruis and Ombaatjie formations, which are characteristically enriched (δ13C = +1 to 342

+8‰ VPDB), but is consistent with the Maieberg Formation (Fig. 4H). 343

In the western syncline (S1, Fig. 6), the upper Ombaatjie, Ghaub and lower 344

Maieberg formations, including the cap dolostone, appear on the hangingwall of the 345

inferred detachment (Section 1, Fig. 7), consistent with westward stratigraphic 346

downcutting analogous to the Ombonde detachment. The estimated thickness of 347

Page 12 of 35



Draft

13

13

missing strata, 450-‐700 m, implies a steeper ramp angle and/or larger displacement 348

compared with the Ombonde detachment. 349

In 2015, we tested a prediction of the detachment hypothesis: an up-‐ramp 350

extension of the detachment in the eastern syncline, near the village of Otjize (Fig. 351

6). On air photos and satellite imagery (Fig. 8), a channel-‐like salient of Tsumeb 352

Subgroup dolostone, ~5.0 km wide in a N-‐S direction, is seen to cut out much of the 353

underlying, strongly-‐cyclical, Ombaatjie Formation. This is evident on both limbs of 354

the syncline. Alternative interpretations of the channel-‐like structure can be 355

discriminated on the ground. If the structure was a (Ghaub) glacial-‐age paleovalley 356

(Fig. 8A), the cap dolostone would occur in the channel axis as well as outside the 357

channel. Up to 85 m of Ghaub-‐age local relief is known elsewhere on the carbonate 358

shelf (Halverson et al. 2002; Hoffman 2011) and more than 400 m occurs at the 359

southern shelf break on Fransfontein Ridge (Hoffman et al. 2014). In those areas, the 360

thickness of the Maieberg Formation as a whole reciprocates with relief on the 361

Ghaub glacial surface (Fig. 8A). Alternatively, if the structure is a post-‐Otavi Group 362

detachment (Fig. 8B), the cap dolostone (Keilberg Member) would be missing in the 363

channel, and both the footwall and hangingwall should contain strata younger than 364

their equivalents in the central syncline. This is because an Ombonde-‐type 365

detachment should ramp up-‐section in an eastward direction (Fig. 5B). 366

Fig. 9 compares measured sections through the upper Abenab and lower Tsumeb 367

subgroups in the eastern syncline. Sections 4 and 6 are from the axial part of the 368

channel-‐like structure on opposite limbs of the syncline (Fig. 6). Sections 3 and 7 are 369

from south of the structure on opposite limbs, and section 5 is from north of the 370

structure in the synclinal hinge zone (Fig. 6). The cap dolostone (Keilberg Member) 371

does not occur within the confines of the channel-‐like structure (Fig. 8), where 372

dolostone grainstone of the upper Maieberg Formation is in fault contact with marly 373

limestone of the lower Ombaatjie Formation. Missing are ~215 m of the upper 374

Ombaatjie Formation, and between 215 and 350 m of the lower Maieberg and 375

Ghaub formations. The total missing interval, 430-‐565 m, compares favourably in 376

thickness with that in the central and western synclines, 450-‐700 m. 377

Page 13 of 35



Draft

14

14

There is no discernible difference in stratigraphic level of the missing interval 378

across the eastern syncline (sections 4 and 6, Fig. 9). Footwall and hangingwall 379

stratigraphies in the eastern syncline are compared with those in the central and 380

western synclines in a longitudinal section of the detachment along the axial zone 381

(Fig. 7). An unfaulted section from east of the detachment is included for 382

comparison (section 8, Fig. 7). From east to west, the detachment ramps down-‐383

section from the lower Ombaatjie to the lower Rasthof formation in the footwall, 384

and from the upper Maieberg to the upper Ombaatjie formation in the hangingwall. 385

Like the Ombonde detachment, the Ombepera detachment ramps down-‐section 386

westward, which is therefore the inferred direction of normal-‐sense displacement. 387

The hypothesized Ombepera detachment is validated by its predicted up-‐ramp 388

extension in the eastern syncline. 389

We cannot determine the magnitude of displacement from the separation of 390

matching stratigraphic cut-‐offs, because the footwall and hangingwall have no 391

exposed cut-‐offs in common (Fig. 7). However, we can calculate a ramp angle of 2.0°, 392

uncorrected for tectonic shortening, given the stratigraphic climbs of ~516 and 393

~528 m in the footwall and hangingwall respectively (Fig. 7), over a horizontal 394

distance in the dip direction of 15 km. Corrected for an estimated 40% shortening 395

due to folding and thrusting, the original ramp angle was 1.3°, slightly steeper than 396

the Ombonde detachment (1.1°). As the footwall and hangingwall of the three 397

synclines have no stratigraphic cut-‐offs in common, the tectonically corrected 398

distance of 25 km between the four sections (Fig. 6) is a minimum bound on the 399

magnitude of displacement of the Ombepera detachment. Inescapably, strata in the 400

hangingwall of the western syncline restore palinspastically to the east of footwall 401

strata in the eastern syncline (Fig. 7). 402

403

Discussion and remaining problems 404

Arc-‐continent collision having been a common process for billions of years 405

(Brown and Ryan 2011), early extensional detachments of the type described here 406

should occur in other orogenic belts, regardless of whether sharp base-‐level fall is 407

Page 14 of 35



Draft

15

15

essential to their origin. Moreover, most examples on existing margins occur on 408

passive margins, so incipient subduction is not required for their development. Yet 409

to our knowledge, few coherent map-‐scale extensional detachments (mass transport 410

complexes) have been described from paleomargins in ancient orogenic belts. 411

Failure to recognize the Ombepera detachment could have led to serious errors in 412

the interpretation of shelf development. The anomalously thin and argillaceous 413

Abenab Subgroup of the western syncline (Fig. 7) could have been interpreted as 414

evidence for a western shelf margin, although no such margin is apparent in the 415

Ombombo or Tsumeb subgroups in the same area. According to the detachment 416

hypothesis, the Ombaatjie Formation is thin in the western syncline because it is 417

truncated from below. Its more terrigenous character, relative to footwall sections 418

in the eastern syncline, relates to its more inboard, not outboard, palinspastically 419

location on the shelf. 420

The anticlinorium between the western and central synclines includes a west-‐421

facing panel in which the apparent thickness of the Rasthof Formation attenuates 422

northward from 350 to 96 m in a strike distance of just 4.4 km (location a, Fig. 6). In 423

the absence of structural disturbance, the attenuation would imply a north-‐facing 424

shelf margin of upper Rasthof age. In light of the detachment hypothesis, the 425

attenuation could represent instead a lateral footwall ramp. Closely-‐spaced 426

measured sections imply top-‐down truncation, consistent with a footwall ramp 427

having an average ramp angle of 3.3°. On the other hand, the detachment could 428

merely track the Rasthof-‐Gruis formation rheological boundary across a primary 429

shelf margin. In this case, sedimentary facies should be distinct from those in the 430

shelf interior. 431

The lateral ramps bounding the Ombepera detachment are just ~5.0 km apart at 432

the base of the Tsumeb Subgroup in the footwall of the eastern syncline (Fig. 8). The 433

narrowness of the detachment recalls the South Makassar Strait mass transport 434

complex (Fig. 1C) and other submarine landslides (Frey-‐Martinez et al. 2005; 435

Masson et al. 2006; Armandita et al. 2015). Narrowness may help to account for 436

their infrequent recognition on ancient continental margins, including the belated 437

recognition of a second detachment in the Kaoko belt. 438

Page 15 of 35



Draft

16

16

However, reconnaissance mapping and air-‐photo interpretation have not 439

revealed any lateral ramp above the Rasthof Formation for 20 km southward or 40 440

km northward of the detachment axis in the central and western synclines. 441

Additional mapping is required to investigate the possibility that channel-‐like lateral 442

ramps develop preferentially where the detachment surface intersects specific 443

stratigraphic intervals. 444

445

Conclusions 446

A second map-‐scale, coherent, very-‐low-‐angle, detachment structure of Ediacaran 447

age has been found in the Kaoko belt of northwestern Namibia. Like the previously 448

described Ombonde detachment, the Ombepera detachment is a brittle structure 449

that predates orogenic shortening, ramps downward stratigraphically to the west, 450

and underwent normal-‐sense displacement of substantial magnitude (>20 km). The 451

detachment is surprisingly narrow (5 km N-‐S), 25% of its displacement, at the level 452

of the Cryogenian-‐Ediacaran boundary. By analogy with the Ombonde detachment, 453

the stratigraphic age of which is tightly constrained, we interpret the Ombepera 454

structure as the extended domain of a coherent submarine slide (mass transport 455

complex), mobilized on the outer slope of a trench-‐foredeep basin during incipient 456

collision between the Congo craton and the Dom Feliciano–Ribeira arc. Accordingly, 457

the average ramp angle of 1.3° for the Ombepera detachment, relative to the host 458

stratigraphy, would have been augmented at the time of failure by a few degrees of 459

incline related to lithospheric flexure. We expect that submarine slides of this type 460

should occur in other orogenic belts regardless of age. Failure to recognize them can 461

lead to misinterpretation of sedimentary facies and stratigraphic relations. 462

463

Acknowledgements 464

PFH first met John Dewey in April 1971 at a Canadian (then Alberta) Society of 465

Petroleum Geology short course on “The New Global Tectonics” at Banff (Alberta), 466

presented by Dewey and the late William R. Dickinson. He met Kevin Burke a few 467

months later at a night club in downtown Toronto. Together, they convinced him 468

Page 16 of 35



Draft

17

17

that, in tectonics, the present is the key to past. Field work was authorized by the 469

Geological Survey of Namibia and supported in 2010-‐14 by the Earth System 470

Evolution Project of the Canadian Institute for Advanced Research (CIFAR). Roy 471

McG. Miller drew our attention to Tom Clifford’s Saturn slide. We thank Stephen T. 472

Johnston, Francis A. Macdonald and William A. Thomas for thoughtful comments on 473

the manuscript, and Andrew Hynes and John W. F. Waldron for constructive 474

reviews. Ali Polat invited the paper. 475

476

References 477

Alves, A., Janasi, V.A., Neto, M.C.C., Heaman, L., and Simonetti, A. 2013. U-‐Pb 478

geochronology of the granite magmatism in the Embu Terrane: Implications for 479

the evolution of the Central Ribeira Belt, SE Brazil. Precambrian Research, 230: 1-‐480

12, doi: 10.1016/j.precamres.2013.01.018. 481

Armandita, C., Morley, C.K., and Rowell, P. 2015. Origin, structural geometry, and 482

development of a giant coherent slide: The South Makassar Strait mass transport 483

complex. Geosphere, 11: 376-‐403, doi: 10.1130/GES01077.1. 484

Bandeira, J., McGee, B., Nogueira, C.R., Collins, A.S., Trindade, R. 2012. 485

Sedimentological and provenance response to Cambrian closure of the Clymene 486

ocean: The upper Alto Paraguay Group, Paraguay belt, Brazil. Gondwana 487

Research, 21: 323-‐340. 488

Brown, D., and Ryan, P.D. (Editors) 2011. Arc-‐Continent Collision. Springer-‐Verlag, 489

Heidelberg (Germany), 493 p. 490

Bugge, T., Befring, S., Belderson, R.H., Eidvin, T., Jansen, E., Kenyon, N.H., Holtedahl, 491

H., and Sejrup, E., 1987. A giant three-‐stage submarine slide off Norway. Geo-‐492

Marine Letters, 7: 191-‐198. 493

Butler, R.W.H., and Paton, D.A. 2010. Evaluating lateral compaction in deepwater 494

fold and thrust belts: How much are we missing from “nature’s sandbox”? GSA 495

Today, 20: 4-‐10, doi: 10.1130/GSATG77A.1. 496

Page 17 of 35



Draft

18

18

Clifford, T.N. 1962. Notes on nappes in the Otavi facies of northern South West 497

Africa. University of Leeds (United Kingdom), Research Institute of African 498

Geology, 6th Annual Report, pp. 44-‐45. 499

Clifford, T.N. 2008. The geology of the Neoproterozoic Swakop-‐Otavi transition zone 500

in the Outjo District, northern Damara Orogen, Namibia. South African Journal of 501

Geology, 111: 117-‐140 and Maps 1-‐3, doi: 10.2113/gssajg.111.1.119. 502

Cloetingh, S.A.P.L., Wortel, M.J.R., and Vlaar, N.J. 1982. Evolution of passive 503

continental margins and initiation of subduction zones. Nature, 297: 139-‐142. 504

Chemale, Jr., F., Mallmann, G., Bitencourt, M.F., and Kawashita, K. 2012. Time 505

constraints on magmatism along the Major Gercino Shear Zone, southern Brazil: 506

Implications for West Gondwana reconstruction. Gondwana Research, 22: 184-‐507

199, doi: 10.1016/j.gr.2011.08.018. 508

Dingle, R.V. 1977. The anatomy of a large submarine slump on a sheared continental 509

margin (SE Africa). Journal of the Geological Society, London, 134: 293-‐310. 510

Dürr, S.B., and Dingeldey, D.P. 1996. The Kaoko belt (Namibia): Part of a late 511

Neoproterozoic contionental-‐scale strike-‐slip system. Geology, 24: 503-‐506. 512

Faleiros, F.M., Campanha, G.A.C., Martins, L., Vlach, S.R.F., and Vasconcelos, P.M. 513

2011. Ediacaran high-‐pressure collision metamorphism and tectonics of the 514

southern Ribeira Belt (SE Brazil): Evidence for terrane accretion and dispersion 515

during Gondwana assembly. Precambrian Research, 189: 263-‐291, doi: 516

10.1016/j.precamres.2011.07.013. 517

Frets, D.C. 1969. Geology and structure of the Huab-‐Welwitschia area, South West 518

Africa. University of Cape Town (South Africa), Precambrian Research Unit, 519

Bulletin 5, 235 p. 520

Frey-‐Martinez, J., Cartwright, J., and Hall, B. 2005. 3D seismic interpretation of slump 521

complexes: examples from the continental margin of Israel. Basin Research, 17: 522

83-‐108, doi: 10.111/j.1365-‐2117.2005.00255.x. 523

Fuck, R.A., Brito Neves, B.B., and Schobbenhaus, C. 2008. Rodinia descendents in 524

South America. Precambrian Research, 160: 108-‐126, doi: 525

10.1016/j.precamres.2007.04.018. 526

Page 18 of 35



Draft

19

19

Ganade de Araujo, C.E., Rubatto, D., Hermann, J., Cordani, U.G., Caby, R., and Basei, 527

M.A.S. 2014. Ediacaran 2,500-‐km-‐long synchronous deep continental subduction 528

in the West Gondwana Orogen. Nature Communications, 5: 5198, 1-‐8, doi: 529

10.1038/ncomms6198. 530

Goscombe, B., Gray, D., Armstrong, R., Foster, D.A., and Vogl, J. 2005. Event 531

geochronology of the Pan-‐African Kaoko Belt, Namibia. Precambrian Research, 532

140: 103.e1-‐103.e41, doi: 10.1016/j.precamres.2005.07.003. 533

Goscombe, B.D., and Gray, D.R. 2007. The Coastal Terrane of the Kaoko Belt, 534

Namibia: Outboard arc-‐terrane and tectonic significance. Precambrian Research, 535

155: 139-‐158, doi: 10.1016/j.precamres.2007.01.008. 536

Goscombe, B.D., and Gray, D.R. 2008. Structure and strain variation at mid-‐crustal 537

levels in a transpressional orogen: A review of Kaoko Belt structure and the 538

character of West Gondwana amalgamation and dispersal. Gondwana Research, 539

13: 45-‐85, doi: 10.1016/j.gr.2007.07.002. 540

Goscombe, B., Hand, M., and Gray, D. 2003. Structure of the Kaoko Belt, Namibia: 541

progressive evolution of a classic transpressional orogen. Journal of Structural 542

Geology, 25: 1049-‐1081. 543

Goscombe, B., Hand, M., Gray, D., and Mawby, J. 2003. The metamorphic architecture 544

of a transpressional orogen: the Kaoko Belt, Namibia. Journal of Petrology, 44: 545

679-‐711. 546

Guj, P., 1970. The Damara mobile belt in the south-‐western Kaokoveld, South West 547

Africa. University of Cape Town (South Africa), Precambrian Research Unit, 548

Bulletin 8, 168 p. 549

Halverson, G.P., Hoffman, P.F., and Schrag, D.P. 2002. A major perturbation of the 550

carbon cycle before the Ghaub glaciation (Neoproterozoic) in Namibia: Prelude to 551

snowball Earth? Geochemistry, Geophysics, Geosystems, 3(6): 1-‐24, doi: 552

10.1029/2001GC000244. 553

Halverson, G.P., Hoffman, P.F., Schrag, D.P., Maloof, A.C., and Rice, A.H.N. 2005. 554

Toward a Neoproterozoic composite carbon-‐isotope record. Geological Society of 555

America Bulletin, 117: 1181-‐1207, doi: 10.1130/B25630.1. 556

Page 19 of 35



Draft

20

20

Hampton, M.A., Lee, H.J., and Locat, J. 1996. Submarine landslides. Reviews of 557

Geophysics, 34: 33-‐59. 558

Heilbron, M., Tupinambá, M., Valeriano, C.M., Armstrong, R., Siva, L.G.E., Melo, R.S., 559

Simonetti, A., Soares, A.C.P., and Machado, N. 2013. The Serra da Bolívia complex: 560

The record of a new Neoproterozoic arc-‐related unit at Ribeira belt. Precambrian 561

Research, 238: 158-‐175, doi: 10.1016/j.precamres.2012.09.014. 562

Hoffman, P.F. 2011. Strange bedfellows: glacial diamictite and cap carbonate from 563

the Marinoan (635 Ma) glaciation in Namibia. Sedimentology, 58: 57-‐119, doi: 564

10.1111/j.1365-‐3091.2010.01206.x. 565

Hoffman, P.F., and Hartz, E.H. 1999. Large, coherent, submarine landslide associated 566

with Pan-‐African foreland flexure. Geology, 27: 687-‐690. 567

Hoffman, P.F., and Halverson, G.P. 2008. Otavi Group of the western Northern 568

Platform, the Eastern Kaoko Zone and the western Northern Margin Zone. In The 569

Geology of Namibia, Volume 2, Neoproterozoic to Lower Palaeozoic. Edited by 570

R.McG. Miller. Geological Survey of Namibia, Windhoek, pp. 13-‐69–13-‐135. 571

Hoffman, P.F., Bellefroid, E., Hodgin, E.B., Johnson, B.W., Kunzmann, M., Sansjofre, P., 572

Strauss, J.V., and Schrag, D.P. 2014. A Cryogenian-‐early Ediacaran carbonate shelf 573

break dominated by glacial paleotopography, Fransfontein Ridge, Kunene Region, 574

Namibia. Geological Society of America, 2014 Annual Meeting Abstracts, Paper 575

130-‐2. 576

Hoffmann, K.-‐H. 1987. Application of tectonostratigraphic terrane analysis in the 577

Damara Province of central and northern Namibia. In Abstract Volume, A.L. Du 578

Toit Golden Jubilee Conference on Tectonostratigraphic Terrane Analysis, pp. 25-‐579

27, Royal Society and Geological Society of South Africa, Cape Town. 580

Hoffmann, K.-‐H., and Prave, A.R. 1996. A preliminary note on a revised subdivision 581

and regional correlation of the Otavi Group based on glacigenic diamictites and 582

associated cap dolostones. Communications of the Geological Survey of Namibia, 583

11: 77-‐82. 584

Jung, S., Brandt, S., Nebel, O., Hellebrand, E., Seth, B., and Jung, C. 2014. The P-‐T-‐t 585

paths of high-‐grade gneisses, Kaoko Belt, Namibia: Constraints from mineral data, 586

Page 20 of 35



Draft

21

21

U-‐Pb allanite and monazite and Sm-‐Nd/Lu-‐Hf garnet ages and garnet ion probe 587

data. Gondwana Research, 25: 775-‐796, doi: 10.1016/j.gr.2013.05.017. 588

Konopásek, J., Košler, J., Sláma, J., and Janoušek, V. 2014. Timing and sources of pre-‐589

collisional Neoproterozoic sedimentation along the SW margin of the Congo 590

craton (Kaoko Belt, NW Namibia). Gondwana Research, 26: 386-‐401, doi: 591

10.1016/j.gr.2013.06.021. 592

Konopásek, J., Košler, J., Tajčmanová, L., Ulirch, S., and Kitt, S. 2008. Neoproterozoic 593

igneous complex emplaced along major tectonic boundary in the Kaoko Belt (NW 594

Namibia): ion probe and LA-‐ICP-‐MS dating of magmatic and metamorphic 595

zircons. Journal of the Geological Society, London, 165: 153-‐165. 596

Konopásek, J., Kröner, S., Kitt, S.L., Passchier, C.W., and Kröner, A. 2005. Oblique 597

collision and evolution of large-‐scale transcurrent shear zones in the Kaoko Belt, 598

NW Namibia. Precambrian Research, 136: 139-‐157. 599

Lehmann, J., Saalmann, K., Naydenov, K.V., Milani, L., Belyanin, G.A., Zwingmann, H., 600

Charlesworth, G., Kinnaird, J.A. 2015. Structural and geochronological constraints 601

on the Pan-‐African tectonic evolution of the northern Damara Belt, Namibia. 602

Tectonics, doi: 10.1002/2015TC003899. 603

Masson, D.G., Harbitz, C.B., Wynn, R.B., Pedersen, G., and Løvholt, F. 2006. Submarine 604

landslides: processes, triggers and hazard prediction. Philosophical Transactions 605

of the Royal Society, London, ser. A, 364: 2009-‐2039, doi: 606

10.1098/rsta.2006.1810. 607

McGee, B., Collins, A.S., Trindade, R.I.F., and Jourdan, F. 2015a. Investigating mid-‐608

Ediacaran glaciation and final Gondwana amalgamation using coupled 609

sedimentology and 40Ar/39Ar detrital muscovite provenance from the Paraguay 610

Belt, Brazil. Sedimentology, 62: 130-‐154, doi: 10.111/sed.12143. 611

McGee, B., Collins, A.S., Trindade, R.I.F., and Payne, J. 2015b. Age and provenance of 612

the Cryogenian to Cambrian passive margin to foreland basin sequence of the 613

northern Paraguay Belt, Brazil. Geological Society of America Bulletin, 127: 76-‐614

86, doi: 10.1130/B30842.1. 615

Page 21 of 35



Draft

22

22

McKenzie, D.P. 1969. Speculations on the consequences and causes of plate motions. 616

Geophysical Journal International (Geophysical Journal of the Royal Astronomical 617

Society), 18: 1-‐32. 618

Miller, R.McG. 2008. The Geology of Namibia, Volume 2, Neoproterozoic to Lower 619

Palaeozoic. Geological Survey of Namibia, Windhoek. 620

Morley, C.K., King, R., Hillis, R., Tingay, M., and Backe, G. 2011. Deepwater fold and 621

thrust belt classification, tectonics, structure and hydrocarbon prospectivity: A 622

review. Earth-‐Science Reviews, 104: 41-‐91. 623

Oyhantçabal, P., Siegesmund, S., Wemmer, K., and Passchier, C.W. 2011. The 624

transpressional connection between Dom Feliciano and Kaoko Belts at 580-‐550 625

Ma. International Journal of Earth Sciences (Geologische Rundschau), 100: 379-‐626

390, doi: 10.1007/s50031-‐010-‐0577-‐3. 627

Oyhantçabal, P., Siegesmund, S., Wemmer, K., Presnyakov, S., and Layer, P. 2009. 628

Geochronological constraints on the evolution of the southern Dom Feliciano Belt 629

(Uruguay). Journal of the Geological Society, London, 166: 1075-‐1084, doi: 630

10.1144/0016-‐76492008-‐122. 631

Rowan, M.G., Peel, F.J., and Vendeville, B.C. 2004. Gravity-‐driven fold belts on passive 632

margins. In Thrust Tectonics and Hydrocarbon Systems. Edited by McClay, K.R. 633

American Association of Petroleum Geologists, Memoir 82, pp. 157-‐182. 634

SACS (South African Committee for Stratigraphy) 1980. Damara sequence. In 635

Stratigraphy of South Africa, Part 1. Edited by Kent, L.E. Handbook of the 636

Geological Survey of South Africa, 8: 415-‐438. 637

Stanistreet, I.G., and Charlesworth, E.G. 1999. Damaran basement-‐cored fold nappes 638

incorporating pre-‐collisional basins, Kaoko Belt, Namibia, and controls on 639

Mesozoic supercontinental breakup. University of Witwatersrand (South Africa), 640

Economic Geology Research Unit, Information Circular 332, 14 p. 641

Stanistreet, I.G., Kukla, P.A., and Henry, G. 1991. Sedimentary basinal responses to a 642

late Precambrian Wilson cycle: The Damara orogen and Nama foreland, Namibia. 643

Journal of African Earth Sciences, 13: 141-‐156. 644

Suppe, J. 1984. Kinematics of arc-‐continent collision, flipping of subduction, and 645

back-‐arc spreading near Taiwan. Geological Society of China, Memoir 6, pp. 21-‐33. 646

Page 22 of 35



Draft

23

23

Tupinambá, M., Heilbron, M., Valeriano, C., Júnior, R.P., Dios, F.B., Machado, N., Silva, 647

L.G.E., Cesar, J., and Almeida, J.C.H. 2012. Juvenile contribution of the 648

Neoproterozoic Rio Negro Magmatic Arc (Ribeira Belt, Brazil): Implications for 649

Western Gondwana amalgamation. Gondwana Research, 21: 422-‐438, doi: 650

10.1016/j.gr.2011.05.012. 651

Will, T.M., Okrusch, M., and Gruner, B.B. 2004. Barrovian and Buchan type 652

metamorphism in the Pan-‐African Kaoko belt, Namibia: implications for its 653

geotectonic position within the framework of Western Gondwana. South African 654

Journal of Geology, 107: 431-‐454. 655

656

Figure captions 657

Fig. 1. Interpretations of seismic reflection profiles (redrawn from Butler and Paton 658

2010; Armandita et al. 2015) showing Late Cretaceous (A) and early Pliocene (B) 659

submarine landslides (mass transport complexes) within fine-‐grained terrigeous 660

and carbonate sediments respectively on the passive continental margins of 661

southwest Africa and southeast Borneo. The scale is the same for each profile and 662

the Borneo margin is geographically ‘reversed’ for kinematic comparison. The 663

Namibian margin faces the South Atlantic Ocean; the Borneo margin (Paternoster 664

Platform) faces the South Makassar Straits Basin and the West Sulawesi thrust-‐fold 665

belt. The slides include extended domains characterized by normal-‐fault duplexes 666

(green lines) and shortened domains by thrust duplexes (blue lines), both soled on a 667

basal detachment (red). Note syn-‐slide growth strata on the Namibian margin. In the 668

South Makassar Strait slide, the strain domains migrated down-‐dip over time, 669

resulting in thrusting succeeded by cospatial normal faulting. Planform of the South 670

Makassar slide mass (C) shows its width to be only a fraction of its maximum 671

translation, unlike detachments of tectonic origin. The barbed line in C indicates the 672

steep frontal ramp in B, not the toe of the detachment. 673

Fig. 2. Cratons (clear), Ediacaran–early Cambrian orogenic belts (shaded), and 674

Mesozoic–Cenozoic belts (hachured) of southwest Gondwana. Dotted lines are the 675

present shorelines of South America, southern Africa and West Antarctica in a pre-‐676

Page 23 of 35



Draft

24

24

drift reconstruction. SF – Sao Francisco, RP – Rio Plata, P – Pampean, PP – 677

Paranápanema, Zam. – Zambesi belt (modified from Fuck et al. 2008). Black stars 678

indicate the Dom Feliciano – Ribeira magmatic arc, which collided with the Congo 679

craton to form the Kaoko belt. 680

Fig. 3. Tectonic map of northwest Namibia showing elements of the Congo and 681

Kalahari cratons, and the Kaoko and Damara orogenic belts (adapted from 682

Hoffmann 1987). Boxes in the Eastern Kaoko zone (EKz) frame the Ombonde (Fig. 683

5) and Ombepera (Fig. 6) detachment areas. Dashed magenta line is the boundary of 684

the Kunene Region. 685

Fig. 4. Stratigraphic column of the Neoproterozoic Damara Supergroup in the 686

Eastern Kaoko zone Namibia (Hoffman and Halverson 2008). (A) Geologic periods, 687

(B) Groups, (C) Subgroups, (D) Formations and Members, (E) radiometric 688

chronology, (F) tectonic phase, (G) generalized lithology, (H) carbonate C-‐isotope 689

composition in per mil relative to the VPDB (Vienna PeeDee Belemnite) standard. 690

Fig. 5. (A) Geologic map of the Grootberg syncline (location, Fig. 3). Red line is the 691

Ombonde detachment (ticks on hangingwall), a brittle, low-‐angle, normal-‐sense 692

detachment that ramps stratigraphically downward from east to west and omits 693

~400 m of stratigraphic section at any location (Hoffman and Hartz 1999). (B) 694

Unfolded W-‐E cross-‐section showing ramp-‐flat geometry of the detachment (red 695

line), matching footwall and hangingwall cutoffs (e.g., arrows b and h, keyed to 696

locations on map), and paleovalley (d) filled by dolomite-‐chert conglomerate (basal 697

Renosterberg Formation) that is incised across the detachment. Minimum age of the 698

detachment is the sub-‐Renosterberg paleovalley (d). Its maximum age is the top of 699

the Tsumeb Subgroup at location (i). (C) Unfolded N-‐S cross-‐section before the 700

detachment showing stratigraphic truncation of the basal Nabis Formation (g), basal 701

Rasthof Formation (i) and basal Gruis Formation (k). Truncations are related to 702

episodic uplift and down-‐to-‐the-‐north rotation of the Makalani dip-‐slope, the 703

footwall of an inferred south-‐dipping, syn-‐rift, normal fault, older and unrelated to 704

the Ombonde detachment. 705

Page 24 of 35



Draft

25

25

Fig. 6. Geologic map of a part of the eastern Kaoko belt (location, Fig. 3), showing 706

the Ombepera detachment (red line with ticks on the hangingwall) exposed in three 707

synclines (S1-‐3). Numbered lines indicate measured stratigraphic sections displayed 708

graphically in Fig. 7 and 9. The detachment widens (N-‐S direction) from the eastern 709

syncline to the central and western synclines. The Ombaatjie Formation (dark 710

green) in the hangingwall of the western syncline restores palinspastically to the 711

east of the same formation in the footwall of the eastern syncline. Abrupt thinning of 712

the Rasthof Formation at location a is tentatively interpreted as a lateral ramp. 713

Fig. 7. Axial (W-‐E) section of the Ombepera detachment (red line), with graphic logs 714

of measured sections of the Ombombo, Abenab and lower Tsumeb subgroups 715

(section 1-‐6 locations, Fig. 6; section 8 located 57 km due East of Section 6). The 716

detachment ramps stratigraphically downward from east to west in both the 717

hangingwall and footwall, and 430-‐700 m of strata are omitted across the 718

detachment in each section. Inset in section 2 shows depleted C-‐isotope values in 719

the hangingwall, diagnostic of the Maieberg Formation (Fig. 4H). Keilberg Member 720

cut-‐off in the hangingwall at magenta x corresponds to footwall cut-‐off at magenta y. 721

Ghaub-‐Keilberg diamictite-‐cap dolostone interval is missing in sections 2, 4 and 6. 722

Field notebook section numbers given alongside columns. 723

Fig. 8. Oblique satellite image (GoogleEarth) looking northwestward at the eastern 724

syncline near the village of Otjize (Fig. 6). The light dotted line is the Ombepera 725

detachment where it cuts out the lower Maieberg and upper Ombaatjie formations. 726

Measured sections are numbered as in Fig. 7 and 9. Insets A and B show 727

downcuttings of different origin: (A) Ghaub glacial-‐age paleovalley and (B) post-‐728

Maieberg extensional detachment. Observed truncation of the lower Maieberg 729

Formation including its basal Keilberg Member favours interpretation B. 730

Fig. 9. Transverse (N-‐S) sections of the Ombepera detachment (red line) on both 731

limbs of the Otjize syncline (S3 in Fig. 6), with graphic logs of measured 732

stratigraphic sections (locations, Fig. 6) of the upper Abenab and lower Tsumeb 733

subgroups (Fig. 4). The upper Ombaatjie and lower Maieberg formations are missing 734

Page 25 of 35



Draft

26

26

in the axial zone of the detachment (sections 4 and 6), compared with sections to the 735

north (section 5) and south (sections 3 and 7) of the detachment. 736

737

Page 26 of 35



Draftex tended domain

shor tened domain

p o s t - s l i d e s e c t i o n

b a s e o f s y n - s l i d e g r o w t h s t ra t a

p o s t - r i f t s e c t i o n

s y n - r i f t s e c t i o n

0

1

2

3

4

5sec.

(tw

o-w

ay tr

avel

tim

e)

approximately 2x vertical exaggeration

10 km

SW NE

redrawn from Butler and Paton (2010)original seismic line: CGGVeritas and the Virtual Seismic Atlas

NWSE1

2

3

4

5

sec.

(tw

o-w

ay tr

avel

tim

e)

redrawn from Armandita et al. (2015)

Namibian continental margin

S outh M ak assar St rait Bas i n, I ndonesia

Late Miocene

Late Pliocene - Pleistocene

f luid escape

shor

tene

d d

omai

n

exte

nded

dom

ain

SE

NW

10 km

Paternoster PlatformWest Sulawesi thrust-fold bel t

10 km

A

B

C

Page 27 of 35



DraftCongo

KalahariAmaz

onia

WAfr

RPP

PP

SF

Damara

Kao

koG

ariep

West

Gondwan

a

East

Afr

ica

n

Zam.

Alto Paraguay

1000 km

Page 28 of 35



Draft18o

20o

12o 14o 16o 18o

22o

24o

100 km

NaukluftAllochthon

Kamanjab inlier

O t a v i f o l d b e l t

Kunen e R.A N G O L A

Kalahar i Desert

Omaheke Desert

C o n g o c r a t o n

Rehoboth Inlier

Windhoek

K

homas

zone

Hakos marg

in

Swakop

te

rrane W

Kz

Etendeka

Witvlei zone

EK

z CK

z

Epupa inlier

Er

Namib Desert

KALAHARI CRATONNeoproterozoic - Cambrian cover

Hakos margin: deformed north-facing margin

1.0 - 1.4 Ga basement: Namaqua Belt

Khomas zone: south-facing accretionary prism

Swakop terrane: arc-bearing microcontinent

Outjo zone: Swakop-Congo collision(?) zoneDAMARA BELT (ca 550 Ma)

CONGO CRATON Neoproterozoic cover

1.8 - 2.0 Ga basement

KAOKO BELT (ca 590 Ma)

Central zone: deformed margin

Western (Coastal) zone: accreted forearc

CKz

WKz

Carb. - Cenozoic cover

SKz Southern zone: deformed deep-sea fan

Eastern zone: foreland thrust-fold beltEKz

geosuture thrust (teeth on hangingwall)

134 Ma igneous rocks

L EGEND plateau

S o u t h

A t l a n t i c

O ce a n

Fig. 5

Fig. 6

ST

PS

Outjo zone

SKz

Page 29 of 35



DraftHuttenberg

Elandshoek

Maieberg

Ombaatjie

Rasthof

Gruis

Okakuyu

Devede

Beesvlakte

Sesfontein

Renosterberg

Tsu

meb

Ab

enab

Om

bo

mb

o

MU

LDE

N

Ghaub

Chuos

Nabis

NO

SIB

ED

IAC

AR

AN

CR

YO

GE

NIA

NTO

NIA

N SY

N-R

IFT

SH

ELF

(P

OS

T-R

IFT

)FO

RE

DE

EP

Keilberg Mb

dolomite

limestone

marlstone

sandstone

siltstone

diamictitecarbonatediamictite

OTAVI

Legend

635

659717

760

~590

746

(Ma)

-5 0 5 10δ13Ccarb (o/oo VPDB)

-10

A B C D E F G

200

0 (m)

H

400 Isotope dataat Ongongo (46 km SSE of Otj ize).

Page 30 of 35



Draft

CenozoicCretaceousMulden Gp Renosterberg Fm

upper TsumebMaieberg FmOmbaatjie FmGruis FmRasthof FmBeesvlaakte FmNabis FmbasementOrosirian

Nosib Group

Ombombo

Abenab Subgp

Tsumeb Subgp

Etendeka Group

Ota

vi G

rou

p

0 2 4 6 8 10 kmOmbonde detachmentoblique reverse faultvehicle track

14o05'E 14o10'E 14o15'E

14o10'E14o05'E

14o15'E

19o30'S

19o35'S

19o40'S

19o45'S

KF KF

20.5 km

West East

North South

3.0 km

0.3 km

paleovalley

paleovalley

Ombonde detachment

Makalani dip-slope

Grootberg syncline

10 km

O mbonde R.

A

B

C

Kalahari Group

a b c

d

e f

b h

i

g

h

j

kig

d

j

k

Page 31 of 35



Draft

7

6

5

4

3

2

12

0 1 2 3

km

N

Se

sfo

nte

in

Th

rust

Okamborombonga

Okaaru

Otji kukutu

Otjo-matemba

Cont ract ional st ruct ures

Ombepera detachment

Graded road C43 O tj i taimo Ri ver

h a n g i n g wa l l

travertine

Ediacaran Tsumeb Subgroup Cryogenian Abenab Subgroup Tonian Ombombo Subgroup

Elandshoek-Huttenberg fms Ombaatjie Formation Okakuyu Formation

Maieberg Formation Gruis Formation Devede Formation

Cryogenian Tsumeb Subgroup Rasthof Formation Beesvlakte Formation

Ghaub (glacial) Formation Chuos (glacial) Formation

“

Geology by P.F. Ho�man & G.P. Halverson

13o35’00”E 13o40’00”E 13o45’00”E

18o45

’00”

S18

o50

’00”

S

Ombepera

Otjize

S1

S2

S3

a

Page 32 of 35



Draft

hangingwall

footwall

Datum = base Rasthof Fm

P210

6

P144

1/42

J141

5E1

414

P144

3

P144

4

100

200

300

0

400

-100

-400

-300

-200

-700

-600

-500

600

500

P650

1

P165

9

P650

4P6

500

700

800

900

1100

1000

1200

-800

P150

22

West East

100

200

300

0

400

-100

-400

-300

-200

-700

-600

-500

600

500

700

800

900

1100

1000

1200

-800

(m)

G11

03

1

62 4

strati�ed polymictic diamictitesiltstone w ice-rafted debrisconglomeratequar tz sandstonequar tz siltstoneargillite

tepeed microbialaminiteooid/intraclast grainstonepeloidal grainstone (cap dolomite)columnar or mounded stromatolitethrombolitic microbialaminiterollup microbialaminiteribbonitemarly ribboniterhythmite (dolomite)rhythmite (limestone)

L E G E N DTerrigenous

Carbonate

massive carbonate diamictitestrati�ed carbonate diamictitemassive polymictic diamictite

Rasthof

Gruis

Ombaatjie

Okakuyu

Devede

Chuos

8

P131

7P2

095

Elandshoek

upperMaieberg

middleMaieberg

middleMaieberg

u. Mbg

GhaubKeilberg

Ghaub

Keilberg

Ombaa- tjie

(m)

x

y

57 km4 km7 km4.5 km

P132

6

-10 -5 0δ13C

Page 33 of 35



Draft Ar

Ab

Tml

Te

Tmu

Tmu

Te

Ar

AbAg Ag

PALEOVALLEY DETACHMENT

N

Otjize

lateral ramp

lateral rampm ove m e n t

d irection

Keilberg MbC43

Ar

Ab

Ac

Ag

Tml

Od

7

6

54

Tml/Tmu - lower/upper Maieberg Fm Ab - Ombaatjie Fm Ag - Gruis Fm Ar - Rasthof Fm Ac - Chuos FmOd - Devede Fm

Ar

Od

Tmu

Tmu

Tmu

A B

Page 34 of 35



Draft

P15024Otj itaimo R.

P6502-3 Okaaru

P9504-7Otjomatemba

P15023Otjikukutu

Ombepera detachment

0

50

100

150

200

250

300

350

400

450

500

0

50

100

150

200

250

300

350

400

450

500

P15022 Otjize

-50

-100

-50

-100

Gruis Fm

Ombaatj ie Fm

M aieberg

Fm

up

per

Mai

eber

g F

m

up

per

Mai

eber

g F

m

SW N SE

Stra

tigr

aphi

c he

ight

(m) r

elat

ive

to th

e ba

se o

f the

Om

baat

jie F

m

3

45

67

Rasthof

Fm

Kei lberg Mb

CW CE

Ghaub Fm

rhythmite (seaf loor cement)

cap swaley dolopelarenite

cap tubestone stromatolite

littoral microbialaminite

intraclast-ooid grainstone

columnar stromatolite

ribbonite

marly ribbonite

limestone rhythmite

carbonate diamictite, rudite

terrigenous sandstone

argillite, terrigenous siltstone

Datum = baseOmbaatj ie Fm

cycle b8

b7

b5-6

b1-4

9 km 6 km 6 km 4 km

Page 35 of 35



Early extensional detachments in a contractional orogen · Draft Early extensional detachments in a contractional orogen: coherent, map-scale, submarine slides (mass transport complexes)

Documents